This proposal requests K07 support for Dr. Elizabeth Burnside to become an independent investigator and leader in the use of computer models to improve the early detection of breast cancer. This award will allow Dr. Burnside to optimize and validate a computer model called a Bayesian network designed to provide decision-support for radiologists interpreting mammography. This project is a multidisciplinary effort integrating computer science, applied medical informatics, population health science and breast imaging. The University of Wisconsin is an ideal environment for this work because it provides access to renowned researchers in these areas and individuals experienced with the inclusion of diverse populations in women's health research. To achieve these research and career goals, the candidate will follow a career development plan that consists of 1) acquisition of advanced research competencies in artificial intelligence, machine learning algorithms, probability theory, Bayesian reasoning, research ethics, and clinical trial design; 2) mentorship from nationally and internationally recognized experts; and 3) innovative research in computer-aided decision support in breast cancer diagnosis. The current state of breast cancer screening reveals a wide variability of practice and a shortage of expert mammographers. Dr. Burnside's model envisions using expert knowledge to improve non- expert performance in mammography. The model complements current computer-assisted detection algorithms by contributing information to crucial post-discovery aspects of mammography: interpretation and decision-making. Using probability theory, the model defines relationships between a patient's demographic risk factors, her mammographic findings, and diseases of the breast to generate a breast cancer risk assessment uniquely tailored to her. If successful, the model will improve the accuracy of decision-making, decrease the human costs of underperformance and variability, and deliver personalized medicine to patients who undergo mammography.